We want to highlight that neon v2.0.0+ has been optimized for much better performance on CPUs by enabling Intel Math Kernel Library (MKL). See the new features in our latest release. We use neon internally at Intel Nervana to solve our customers' problems across many

Training VGG with 16-bit floating point on 1 Titan X takes ~10 days (original paper: 4 GPUs for 2-3 weeks).

2.5s/macrobatch (3072 images) on AlexNet on Titan X (Full run on 1 GPU ~ 26 hrs).

Swappable hardware backends: write code once and then deploy on CPUs, GPUs, or Nervana hardwareįor fast iteration and model exploration, neon has the fastest performance among deep learning libraries (2x speed of cuDNNv4, see benchmarks).

Model Zoo contains pre-trained weights and example scripts for start-of-the-art models, including: VGG, Reinforcement learning, Deep Residual Networks, Image Captioning, Sentiment analysis, and more.

Support for commonly used layers: convolution, RNN, LSTM, GRU, BatchNorm, and more.

Tutorials and iPython notebooks to get users started with using neon for deep learning.

Designed for ease-of-use and extensibility. Neon is Intel's reference deep learning framework committed to best performance on all hardware. If you have an ongoing need to use this project, are interested in independently developing it, or would like to maintain patches for the community, please create your own fork of the project. Patches to this project are no longer accepted by Intel. Intel will not provide or guarantee development of or support for this project, including but not limited to, maintenance, bug fixes, new releases or updates. This project will no longer be maintained by Intel.

Tabs how to get neon